Import data (Mainly focus on publisher and their console)

## 
## -- Column specification --------------------------------------------------------
## cols(
##   title = col_character(),
##   console = col_character(),
##   genre = col_character(),
##   release_date = col_date(format = ""),
##   publisher = col_character(),
##   developer = col_character(),
##   total_sale = col_double(),
##   na_sale = col_double(),
##   pal_sale = col_double(),
##   japan_sale = col_double(),
##   other_sale = col_double()
## )

Plot for year and sale, Plot for year and console

year of game vs sale

## `summarise()` ungrouping output (override with `.groups` argument)

test for the normality

## 
##  Shapiro-Wilk normality test
## 
## data:  sales
## W = 0.86719, p-value = 8.884e-05

year vs console

Publisher (TOP 10)

## `summarise()` ungrouping output (override with `.groups` argument)
## Adding missing grouping variables: `title`

Select Period 2005-2015

## `summarise()` ungrouping output (override with `.groups` argument)
## Adding missing grouping variables: `title`

Focus on Nitendo

Nintendo Console

console time
Game & Watch 1980-1991
NES 1983-1995
Game Boy (GB) 1989-2003
SNES 1990-1998
Virtual Boy 1995-1996
N64 1996-2003
GameCube (GC) 2001-2007
Game Boy Advance 2001-2008
DS 2004-2014
Wii 2006-2013
Nintendo 3DS (3DS) 2011-
Will U (WiiU) 2012-2016
NES 2016-2017
Switch (NS) 2017-

games on each Nintendo Console during 2005-2015

console n
3DS 59
DS 128
GBA 22
GC 15
Wii 64
WiiU 27
## Analysis of Variance Table
## 
## Response: nin
##              Df  Sum Sq Mean Sq F value   Pr(>F)   
## factor(ind)   5   763.3  152.65   3.211 0.007667 **
## Residuals   309 14689.8   47.54                    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
##  Pairwise comparisons using t tests with pooled SD 
## 
## data:  new_data$nin and new_data$ind 
## 
##   1     2     3     4     5    
## 2 1.000 -     -     -     -    
## 3 0.204 0.139 -     -     -    
## 4 1.000 1.000 0.048 -     -    
## 5 1.000 1.000 0.031 1.000 -    
## 6 1.000 1.000 0.172 1.000 1.000
## 
## P value adjustment method: bonferroni
##   Tukey multiple comparisons of means
##     95% family-wise confidence level
## 
## Fit: aov(formula = nin ~ factor(ind), data = new_data)
## 
## $`factor(ind)`
##            diff         lwr        upr     p adj
## 2-1  0.66686755  -3.9272530  5.2609880 0.9983927
## 3-1  3.92594329  -0.6114860  8.4633726 0.1328747
## 4-1 -1.13250842  -6.8114299  4.5464130 0.9927760
## 5-1  0.64547454  -3.5418828  4.8328319 0.9978559
## 6-1 -1.10629630  -7.4736383  5.2610457 0.9962123
## 3-2  3.25907574  -0.3095221  6.8276736 0.0957558
## 4-2 -1.79937596  -6.7386851  3.1399332 0.9023656
## 5-2 -0.02139301  -3.1327573  3.0899712 1.0000000
## 6-2 -1.77316384  -7.4906519  3.9443243 0.9488801
## 4-3 -5.05845170  -9.9450759 -0.1718275 0.0376380
## 5-3 -3.28046875  -6.3074985 -0.2534390 0.0249517
## 6-3 -5.03223958 -10.7042755  0.6397963 0.1147198
## 5-4  1.77798295  -2.7854361  6.3414020 0.8740112
## 6-4  0.02621212  -6.5945035  6.6469278 1.0000000
## 6-5 -1.75177083  -7.1478517  3.6443100 0.9382788